Skip to content

Latest commit

 

History

History
119 lines (91 loc) · 3.23 KB

cli-reference.md

File metadata and controls

119 lines (91 loc) · 3.23 KB

CLI Reference

The current CLI surface is runtime-first. maple manages diagnostics, PALM lifecycle, worldlines, commitments, provenance, governance, and the local agent demo. palm manages specs, deployments, instances, health, events, and playground backends.

maple command groups

These are the command groups currently exposed by maple --help:

  • version
  • validate
  • ual
  • doctor
  • daemon
  • agent
  • worldline
  • commit
  • provenance
  • financial
  • gov
  • kernel
  • palm

Notes

  • validate currently behaves as a developer stub, not as a finished manifest-validation UX.
  • palm forwards directly to the PALM CLI.
  • Product-style package, model, and fleet commands such as maple build, maple model, and maple up are not currently exposed here.

Common maple flows

# Diagnose local setup
cargo run -p maple-cli -- doctor --model llama3.2:3b

# Start or inspect the daemon
cargo run -p maple-cli -- daemon start --foreground
cargo run -p maple-cli -- daemon status

# Exercise the local agent demo
cargo run -p maple-cli -- agent demo --prompt "log current runtime status"
cargo run -p maple-cli -- agent demo --dangerous --with-commitment --amount 500

# Work with worldlines and provenance
cargo run -p maple-cli -- worldline create --profile agent --label demo-agent
cargo run -p maple-cli -- worldline list
cargo run -p maple-cli -- provenance worldline-history <worldline-id>

# Inspect the runtime
cargo run -p maple-cli -- kernel status
cargo run -p maple-cli -- kernel metrics

palm command groups

These are the command groups currently exposed by palm --help:

  • spec
  • deployment
  • instance
  • state
  • health
  • events
  • playground
  • config
  • status

Common palm flows

# Check daemon connectivity
cargo run -p palm -- status

# Inspect playground backends and set the active one
cargo run -p palm -- playground backends
cargo run -p palm -- playground set-backend \
  --kind local_llama \
  --model llama3.2:3b \
  --endpoint http://127.0.0.1:11434
cargo run -p palm -- playground infer "Summarize current runtime health"

# Inspect rollout state
cargo run -p palm -- spec list
cargo run -p palm -- deployment list

Environment variables

These variables are part of the current surfaced toolchain:

Variable Used by Purpose
PALM_ENDPOINT maple, palm Default API endpoint for CLI requests
OLLAMA_HOST maple doctor Local Ollama endpoint for connectivity checks
PALM_CONFIG palm, palmd Config file path
PALM_PLATFORM palm, palmd Active platform profile
PALM_LISTEN_ADDR palmd Daemon listen address
PALM_LOG_LEVEL palmd Daemon log level
PALM_LOG_JSON palmd Enable JSON logs
PALM_STORAGE_TYPE maple daemon, maple doctor Storage mode override
PALM_STORAGE_URL maple doctor and PALM storage config PostgreSQL connection URL

Config

PALM CLI reads:

~/.config/palm/config.toml

The current fields are:

endpoint = "http://127.0.0.1:8080"
default_platform = "development"
default_namespace = "mapleai"
timeout_seconds = 30

Use explicit flags for one-off overrides and the PALM config file for stable operator defaults.